[1] 0.7470924 2.3672866 0.3287428
University of British Columbia
October 3, 2024
Computer experiments that involve creating data using pseudo-random sampling.
R.Following earlier example, we know some fixed results (like ground truth).
Then, it is possible to use it for accessing performances since “truth” is available.
Furthermore
parameter/s of interest.many parameters for various options.Evaluation of new methods and comparison of alternative methodsWhat’s the norm now?
Reviewed 100 articles taken from Volume 34 of Statistics in Medicine (2015), that included at least 1 simulation study.
No summary plot given, subtle derivation of classifying those reviewed papers using their procedure ADEMP:
Often, want to compare methods where some or all have shown to work, ignoring fact that methods were designed to address slightly different problems.
Usually spend more time here than other steps in ADEMP
Choice of data-generating process
Also, could depend on aims to investigate mehod under:
What people usually do
Varying sample size \(n\) of simulated data is typically used, because performance varies over \(n\).
From Author’s review only 3/100 used resampling methods, rest used some form of parametric model.
When comparing several methods in order to identify the best:
Thanks!